Fractional CRM & Lifecycle Architect · 14 years · ex-Uber Eats EMEA
More revenue from the customers you already have.
Campaigns go out, but it is hard to tell what to change next. I take ownership of the CRM and lifecycle system — economics, architecture, behaviour — and measure each gain against a holdout wherever the volume allows.
In the Uber Eats EMEA team I contributed to CRM for ~250,000 merchants across 17 countries.
Retainers €2,500–€14,000/mo →Hiring for a role? →
He combines deep CRM systems expertise with a very pragmatic, business-first mindset.
Built lifecycle & MarTech systems at
The work so far
What I moved at Uber.
Every number here opens its own ⓘ — how it was measured, and what it does not claim.
The Uber Eats merchant CRM layer in EMEA — owned end to end.
Every merchant message Uber Eats sent in EMEA — 2M a month, 8 languages — ran through systems I owned. Strategy, SQL, templates, deliverability, experimentation — the whole chain.
businesses under CRM ownership
- Markets
- 17 countries
- Languages
- 8
- Comms
- 2M/mo
- Automations
- ~2,000
- Tempo
- 30–50 campaigns/wk
- Cross-org bridge
- 5 teams
CTR uplift across core EMEA lifecycle. 1.53% → 2.28% via systematic A/B testing, AMPscript fixes, AI personalization.
How this is measured
1.53% → 2.28% click-through across the core EMEA lifecycle portfolio — a portfolio-wide shift, not one hero campaign. Source: SFMC reporting, with holdout groups wherever volume allowed a clean control.
Diagnosed and fixed a global SFMC defect — took deliverability from ~84% to 95%+ worldwide, not just EMEA. How I found it →
How this is measured
~84% → 95%+ delivered, verified in raw delivery logs. Root cause was a platform-level SFMC defect I diagnosed and escalated; the fix shipped worldwide, so every Uber market benefits — not just mine.
Selected work
Three proofs. One owner.
A global production case, a signature automation from earlier in my career, and a published strategy memo. Together they answer the real question: how I think before a system exists, and how I operate once it has to scale.
Uber EMEA — Lifecycle Architecture
250,000 merchants. 17 countries. 8 languages. Rebuilt the CRM architecture, fixed a global deliverability defect, co‑authored the FY26 strategy with a €25M forecast.
PrivatBank — Mastercard VIP train trigger
Event-driven, multi-channel automation replacing a manual sales motion: 3× conversion, with concierge validation through a taxi API.
talabat — Subscription & Churn Engine
Habit-loop framework to grow tPro adoption. 2×2 Risk × Value matrix to prevent churn without burning margin.
Where to start
Small first. Deeper only if it pays.
Prices are public and flat. Each step is useful on its own, and you can stop after any of them.
-
01 · Outside€600
Lifecycle Teardown →
Five business days, zero calls. I walk what is visible from the outside and show where the lifecycle tears.
Credited against the Diagnostic within 60 days.
-
02 · Inside€2,900
Diagnostic →
Two weeks inside your stack. An executive map of what leaks revenue and a 90-day operating plan — yours to keep either way.
-
03 · Ongoingfrom €2,500/mo
Retainer →
Ownership of the system as the work continues. Three months first, then month to month.
- Advisor€2,500/mo
- Embedded Architect€6,500/mo
- Fractional Lead€14,000/mo
How the work runs
Four things I keep.
-
01Start
Three months first. Then month to month. A lifecycle system needs one full cycle before its numbers mean anything.
-
02Proof
A holdout wherever the volume allows. Each gain is read against a control group, not against last month’s report.
-
03Focus
No more than three clients. Each system I take on gets a real owner, not a share of a crowded calendar.
-
04Scope
What I don’t do. Paid acquisition, greenfield CRM builds, production-volume QA. If that is the need, I’ll point you somewhere good. Who I can’t help →
If the model fits,let’s talk.
30 minutes, video, no agenda template. I want to understand where revenue is leaking: data, lifecycle, governance, tooling, or ownership. If the problem fits my shape, we agree on Diagnostic or retainer next steps in the same call. If it doesn’t, I’ll say so.